Data Analytics — A New Angle Of Making Your Data Analytics A Great Investment

There is always confusion or stereotype about what data analytics is and how to achieve a good return on analytics. Whether you are in or outside the data or technology industry, you definitely tend to link data analytics with machine learning, artificial intelligence, k-mean clustering, business intelligence, or big data, all these buzzwords. Nonetheless, how to manage it, how you should treat it, and how you should shift your mindset to retrieve the most out of insights benefited by data analytics, are never spoke of. People put too much spotlight on the engineering, mathematical, or technical part of it but seem oblivious to its other equally imperative part, if not more.

There are a few concepts that are definitely not in the scope of data analytics. Data entry, a conventional means of entering records into systems, is just a part of data analytics. It is one of the many possible starting points for a data analytic journey. Probably to your surprise, plotting graphs or looking for patterns from the massive flood of data is not equivalent to data analytics as well. It is because visualization or pattern recognition is just a way of constructing a simple mathematical model to explain the data, which is just another part of data analytics. Data savvies have a tendency of defining or interpreting data analytics as a series of highly technical activities, such as performing neural networks on a set of problems or running multi-variable regression on predicting the capacity of inventory of a retail shop. This interpretation is not all wrong, but it barely reveals the true colors of the power of data analytics.

Data analytics must be an iterative & incremental process and experiment. To achieve an actionable and insightful data analytics investment, these are the components that you should not miss

1. Set a hypothesis to a prioritized business use case that is aligned with the direction of your organization or line of business.

2. Identify relevant data and capture it from unbiased sources

3. Cleanse, transform, enrich your raw data, turning it into a meaningful data set

4. Apply different techniques (descriptive analysis, predictive analysis, or artificial intelligence) to verify your hypothesis

5. Complement technical knowledge with business & human domain knowledge to confirm your conclusion

6. Turn your conclusion into actionable insights

7. Drive business adoption using your actionable insights

The consideration is more than the technical aspect. Doubtlessly, focusing solely on the technical part jeopardizes your analytics investment.

Agility must be lived and breathed in any data analytics projects. That means data analytics should be an iterative and incremental process and such process does not have the luxury of experiencing a lengthy traditional waterfall (plan, design, develop and test). Having a waterfall mindset will kill your endeavor and set you to fail on the first day. It is not uncommon that in the end, it turns out that you think actionable insights are not actionable nor insights at all to the business. Showing your minimal viable product to business and having their feedback on the insights as earlier as possible prevent your analysis result from being covering with dusk.

Technology competency is imperative. It is a technical component and layer that enables storing, processing, computing, analyzing data. In the era of big data (volume, velocity, variety, and veracity), we need greater computational power and infrastructure to analyze and access it; we need tools that can help us put tremendous richness onto the data set; we need to explore structural, semi-structural and un-structural data. Without all these, we are losing a competitive advantage in this digital disruption age. IT cannot hinder analytic development and speed.

To obtain an impactful result, you must bridge the gap between IT/Data and Business. The goal of data analytics is to solve business problems or create actionable insights. It should be business backward but not data forward. Only by leveraging business domain knowledge from business and technical knowledge from data experts will your data analytics be rewarding, successful, impactful.

We should look at data analytics from a new angle.